Vacuum Techniques for Holding Hardwood Parts

Here's a collection of power-user tips for setting up your vacuum table to hardwood stock down to the CNC table when machining. October 16, 2012

Question
I have a 4x8 CNC router with a 15hp Regen vac pump. It does great for processing sheet good cabinet parts. I don't lose parts unless I am at the extreme end of tool life, or the spoilboard is in need of resurfacing. I have experienced a good bit of difficulty trying to hold hardwood board stock to route out corbels, feet, and other odd parts without resorting to screwing the lumber to my spoil board. I have a prospective client that wants a bunch of hardwood parts, about 28"x5" (the geometry of the part fits roughly within this rectangular dimension.)

I am looking for ideas on how to hold it without screwing or clamping. Will a regen pump/blower pull enough vac if I make a dedicated spoilboard out of melamine and focus all 12-15" of Hg to the board the part will be machined out of? I have tried processing these sorts of things in multiple passes, with a cleanup cut and onion skin in the past with not real good results. Does anyone have any other thoughts or ideas to process a bunch of these without manually fixing each one in place?

Forum Responses
(CNC Forum)
From contributor J:
I use 2" wide double sided carpet tape. Liberally apply it to the bottom of the boards and cover the rest of your table with laminate or melamine to minimize vacuum loss. I cut most thick hardwoods with an Onsrud 3/4" chipper/finisher and leave a 1/16" skin to hold everything together. After it's done cutting, I flip the now perforated board upside down and use more carpet tape to stick it to a chunk of melamine, then run it through the widebelt sander to remove the skin. My finished parts are left stuck to the melamine. I've carved some pretty awkward looking solid wood parts this way with good results.

---

From contributor D:
Also using a 3/4" 3 flute downspiral hogger to remove most stock first, say 1/8" before reaching final dimension is a good strategy. It imparts probably the least amount of force onto a part being cut that I can think of, and a good bit of that is pushing down, instead of pulling up away from your vacuum. Grain orientation should be respected if you can modify your programs instead of using the automatically generated toolpaths. Generally, the closer the angle of attack on grain is to 90 degrees (a crosscut), then a conventional cut imparts less force.

The closer the toolpath is to being parallel to grain, a climbing cut imparts less force. This assumes you're removing half or less stock than the tool diameter. If you hit a small chunk of oak going full speed and it might hit you full speed, in the groin. In my experience carving, the tool is not constantly buried, as it maneuvers around the part it cools somewhat, and most chips fall around the part (on a pod and rail machine they fall down into the trough). I machine small parts all day every day, and I bet my bits stay sharp as long or longer than anyone cutting MDF all day. I respect chipload on finishing passes, usually only removing 1/32". Better Vacuum Cups help a lot. They are made with a very grippy hard rubber, the kind I've only felt elsewhere on rock climbing shoes. Pods which only provide an air seal, but pull down to a slick polycarbonate or phenolic block, offer little friction to combat side forces. I cut large tenons and raised panels with cutters in excess of 6" diameter without throwing a part- same feeds with Schmalz pods and that part is a five pound airborne hardwood frisbee.

---

From contributor F:
I want to try this vacuum cover seal available from Thermwood. You drill holes and it concentrates the vacuum where you need it. Watch the video below and about halfway through a door is machined without a panel using this seal as hold down.

---

From contributor M:
I use 1/16" neoprene gasketing, adhesive on one side. Outline on your spoilboard the part and drill small holes through the sheet and into the dado portion of your table to allow free flow. Cover all unused areas with scrap material to maximize the vacuum. When you’re done roll up a little ball of the neoprene and shove in the holes and spoilboard is good as new. You may have to throttle back your feed speeds but I have had good success this way.

---

From the original questioner:
Thanks for the input. I think the thing that concerns me the most is the pump. When I did this for a former employer, we had a true vacuum pump, and I did some of what has been suggested here with good results. I am just concerned that this regen doesn't have enough suck force. Just have to try it out and toss a few parts.

---

From contributor N:
Every part that is run on my machine is either held down by hot glue to an MDF table or when I can get away with it, aluminum nails.

---

From contributor M:
I have them same setup with a 15 hp regenerative blower. I use the method of drilling through the spoiboard and 1/16 gasket with good results. Without knowing the part details, bandsawing the excess off and then profiling can also be a good strategy to reduce torque on the part.

---

From contributor I:
The easiest thing to do is make a dedicated vacuum board the size of the zones you intend to use on the machine. Use melamine MDF and surface the area for the parts to sit on the raw core, route out a simple shallow grid under the part and drill a few holes all the way through in the groves. This allows the full power of the vacuum to pull the parts and have enough flow at the part to make for leaks and less than perfectly flat parts. Unless the wood is really rough or irregular you shouldn't need a gasket.

If you’re not cutting the perimeter you can make the pocket for the part the same size, dog bone the corners and then recess the part in the spoilboard a little to help it from twisting and let you cut smaller parts faster and more accurately.

---

From contributor D:
Forgot the obvious - pneumatic clamps. For really tiny parts this is the best thing. Sure, you can't get 360 degrees around the part, so programs will be two step ordeals. One tip is to leave a referencing tab on the first step, so when you flip your part 180 degrees you have a precise way of locating it since most referencing popup pins will likely be farther apart than the size of your small part. I'm running parts like this right now, a raised panel that is only 5" x 5".

---

From contributor R:
3M spray77 adhesive is an essential tool for this. Also, onion skinning and taking shallow cuts helps.

---

From contributor B:
We did a lot of research into this in the past year. We cut a fair amount of 1 3/4" teak. We have been holding the blanks down with screws into a sheet of 3/4 ply and it is a hassle. We tried other methods w/o success. We decided to try a grid table. I bought a 1' x 2' piece of 3/4 phenolic (garolite xx) to test the concept.

I flycut both sides, cut a grid of 1/4 wide by 3/16 deep grooves leaving 1" squares in between. I drilled 5/8 deep 1/4 holes in the center of each square and around the perimeter, outside the grid. I then drilled and tapped for an quick release fitting on the edge, and got some round rubber gasket. We epoxied this 1x2 piece onto a 2x4 piece of prelam. This allowed the router’s vac pump to hold it fast to the table. We hooked up our other pump that we use for vacuum bagging to the grid table. After positioning a 6"x12" piece on the grid and sucking it down, I proceeded to hit it sideways on the edge with a two pound mallet. I could not get it to move! We've cut a bunch of small parts with this setup and it works like a charm.

We just ordered a 36 x 48 piece of phenolic to make a full size 18 x 96 grid. Our blanks are always rectangular planks. We will be able to cut anything we need on this size table. We leave an 1/8 skin to route off, but that's no big deal to us. I'd rather store the parts in the plank until needed anyway just to protect them from damage.

I do believe we could use a little venturi pump running on compressed air, or a shop vac to hold the parts down with no problem. For several years, I turned bowls of 18", 20", and 24" diameter on a lathe holding them on a form using a shop vac, and later a venturi, after the shop vac died.